The pros and cons of the fourth revision of thalassaemia screening programme in Iran

Abstract

Objective

To evaluate the repercussions of recent changes to the cut-offs used in the first screening step of the pre-marital screening programme for thalassaemia prevention in Iran.

Methods

The profiles of 984 subjects referred to a genetic laboratory, and the tests of 242 parents of children with thalassaemia major were assessed for red blood cell (RBC) indices, haemoglobin (Hb) A2 levels and results of Hb electrophoresis.

Results

Of 407 suspected thalassaemia minor (STM) cases, 18 proved positive for thalassaemia minor on molecular analysis (18/407, confidence interval 2.6–6.9%). If the revised screening cut-offs had been used to determine who would undergo molecular analysis, two of these cases would not have been identified. Only 4.4% of suspected cases with lower than normal RBC indices (mean corpuscular volume <80 fl and mean corpuscular Hb <27 pg) and HbA2 (<3.5%) were diagnosed with thalassaemia minor.

Conclusion

The thalassaemia major prevention programme is performed in two separate steps. One step involves the screening of subjects and identification of β-thalassaemia minor, suspected cases for thalassaemia minor (STM), and normal subject groups. The other step concerns the identification of thalassaemia minor in the STM group. Changing the cut-offs at the first screening step does not result in significant improvement from an economic view, and is associated with significant risk at the second screening step.

Keywords

Beta-thalassaemia pre-marital screening red blood cell indices prenatal diagnosis

Introduction

β-Thalassaemia major is the most common hereditary disease in Iran. The pre-marital screening programme for β-thalassaemia major prevention, which started in 1995¹ is one of the most successful health programmes in Iran, resulting in a significantly reduced incidence of new cases over the last two decades.^2,3 Many new cases are associated with marriages that began before the implementation of the screening programme; others involve particular cultural and economic conditions that have caused difficulties for the programme.^3,4 The screening programme is based on confirming the existence of thalassaemia in couples, or confirming the absence of thalassaemia in at least one partner.¹ If one partner has normal red blood cell (RBC) indices, the couple is not required to follow through with the programme, regardless of the second partner’s condition. RBC indices below the cut-off point (mean corpuscular volume (MCV) <80 fl and mean corpuscular haemoglobin (MCH) <27 pg) indicate that haemoglobin A2 (HbA2) levels should be measured. HbA2 levels above 3.5% in both partners strongly indicate thalassaemia, and the couple is then referred for consultation. However, where HbA2 is lower than 3.5% for both partners (or for one subject whose partner is already diagnosed with β-thalassaemia minor), and where there is no iron deficiency, thalassaemia minor is suspected, and must be determined by molecular tests (Table 1). Despite the cost-benefit and success of this plan, many suspected thalassaemia minor (STM) subjects were referred to genetic laboratories, and issues arose relating to the cost of molecular testing and the time required for final determination (sometimes several months).^5,6

Table 1.

Cut-off details of the thalassaemia prevention programme in Iran (fourth edition).

	Old approach	New approach
Normal^a	MCV ≥ 80 fl and MCH ≥ 27 pg,	MCV ≥ 80 fl and MCH ≥ 27 pg,
β-Thalassaemia minor	HbA2 > 3.5%^b	HbA2 > 3.5%
Suspected for β-thalassaemia minor	MCV < 80 fl or MCH < 27 pg and HbA2 < 3.5%	Low risk: MCV ≥ 75 fl and MCH ≥ 26 pg and HbA2 < 3.2% and HbF < 3%^c
Suspected for β-thalassaemia minor	MCV < 80 fl or MCH < 27 pg and HbA2 < 3.5%	High risk: any other condition

If MCV ≥ 80 fl and MCH ≥ 27 pg, there is no need to check HbA2.

HbA2 must be checked by column chromatography and should only be checked if MCV and MCH are lower than 80 fl and 27 pg, respectively.

HbF must be checked by capillary electrophoresis.

To address these problems, a fourth revision of the thalassaemia screening programme was issued in September 2014 (Table 1; Figure 1), with changes intended to expedite the determination of the state of subjects, and reduce the number of cases referred for molecular examinations. Under the fourth revision, people with RBC indices close to normal are classified as having “low risk for thalassaemia minor” (LRTM) (see Table 1). If one subject (regardless of the condition of their partner) is identified as LRTM, the couple is not required to undergo molecular evaluation. The cut-off levels of MCV and MCH for the LRTM group are reduced to 75 fl and 26 pg, respectively (subjects in the LRTM group must have MCV ≥75 fl and MCH ≥26 pg). However, there is no scientific consensus on the appropriate cut-off points of RBC indices when performing thalassaemia screening, and lower cut-off points have been suggested.^7–10 Although selecting lower cut-offs for screening reduces administrative problems, it has been unclear whether it leads to missed cases of thalassaemia. We therefore aimed to evaluate the risks of missed cases along with the reduction of cut-offs in RBC indices, as outlined by the fourth revision of the β-thalassaemia prevention programme.

Figure 1.

Flow chart for thalassaemia screening (fourth revision). Low risk for thalassaemia minor (LRTM) (MCV ≥ 75 pg and MCH ≥ 27 pg and HbA2 < 3.2% and HbF < 3%), STM: suspected for thalassaemia minor (MCV < 75 fl, or MCH < 27 pg, or 3.5% > HbA2 ≥ 3.2%, or HbF < 3%), subjects with HbA2 ≥ 7% should be referred to haematologist for further evaluation.

Methods

We carried out a retrospective analysis of 984 subjects from Isfahan province, Iran, referred to the genetic laboratory for detection of any β-globin gene mutations. These included 271 couples referred from a thalassaemia prevention centre for pre-marital screening because of low RBC indices (STM group), 54 subjects by personal request and 194 couples with a known case of thalassaemia and haemoglobinopathies referred for the first step of prenatal diagnosis. We also added the haematologic data of 142 thalassaemia major parents as a random thalassaemia minor group that was not included in any screening programme (married before starting pre-marital thalassaemia screening programme).

STM group (n = 407)

All the subjects in this group were referred to the reference genetic laboratory in Isfahan city during 2011–2013 due to having lower than normal RBC indices and normal HbA2 levels (according to the third revision of the thalassaemia prevention programme), in order to determine the presence or absence of β-thalassaemia using molecular evaluation. RBC indices and HbA2 levels were extracted from files. Although not included in the screening programme, the genetic laboratory routinely performed Hb electrophoresis assay for STM subjects before gene analysis. Therefore, Hb electrophoresis data have been included in this study. This group covers 271 couples. In 135 couples, one of the partners had thalassaemia minor and the other one was STM. Both partners in each of the remaining 136 couples were STM.

Personal request group (n = 54)

These subjects were referred to the reference laboratory in Isfahan city during 2011–2013 due to personal requests (not from a prevention programme).

Thalassaemia minor and haemoglobinopathies group (n = 388)

The subjects in this group were found to have thalassaemia minor or haemoglobinopathies (according to HbA2 levels and Hb electrophoresis), and were referred for beta globin gene analysis at the reference laboratory in Isfahan city as the first step for prenatal diagnosis. These included 191 couples (382 subjects) previously diagnosed with thalassaemia minor, and three couples (six subjects) with specified haemoglobinopathy (confirmed genetically).

Parents of patients with thalassaemia major (n = 218)

This category consisted of the parents of 141 thalassaemia major and thalassaemia intermedia patients in one of the thalassaemia treatment centres in Lordegan, a city west of Isfahan provenance (data from some parents were not available). RBC indices, HbA2 levels and Hb electrophoresis of parents were extracted from their files. Information about the subjects, such as age, sex, RBC indices, Hb electrophoresis, HbA2 levels and data of the molecular tests were also examined.

We analyzed the STM group to assess how the changes in threshold RBC indices for the implementation of the molecular tests between the third (MCV >80 fl and MCH >27 pg) and fourth (MCV >75 fl and MCH >26 pg) revisions of the β-thalassaemia prevention programme would affect the probability of missing a β-thalassaemia diagnosis. We analyzed the other groups to assess the distribution of RBC indices associated with β-thalassaemia minor. Differences between groups were evaluated by using independent sample t tests, Pearson chi-square tests and 95% confidence interval (CI) for proportion.

Results

STM group

According to molecular analysis of the 407 STM subjects, 18 (4.4%) had β-thalassaemia minor (CI: 2.6–6.9%), and 28 had haemoglobinopathies. There was one case of HbD-thalassaemia (IVSI-6/HbD), one case homozygous for HbD and one case with the genotype of IVSII-16/IVSII-16 and normal Hb levels (>12 g/dl) (Table 2). STM cases later diagnosed as β-thalassaemia minor had significantly lower mean MCV and MCH than the rest of the STM group (mean for MCV = 72.68 ± 6.17 fl, MCH = 23.2 ± 2.31 pg for minor group, and 77.14 ± 4.8 fl and 25.08 ± 1.8 pg for STM, p value = 0.000 and p value = 0.008, respectively).

Table 2.

Variations of Hb disorders and RBC indices in 46 subjects of the “suspected for thalassaemia” (STM) group.

Haemoglobin disorder	Subjects (N)	MCV(fl)			MCH (pg)			MCV ≥ 75 fl and MCH ≥ 26 pg
Haemoglobin disorder	Subjects (N)	MCV ≥ 80	80 > MCV ≥ 75	75 > MCV	MCH ≥ 27	27 > MCH ≥ 26	MCH < 26	MCV ≥ 75 fl and MCH ≥ 26 pg
HbD heterozygote	13	5	7	1	4	6	3	10
Sickle trait	12	4	5	3	4	5	3	9
β-Thalassaemia minor	18	3	4	11	1	1	16	2
Other Hb disorders	3	3	0	0	0	2	1	3
Total	46	15	16	15	9	14	23	24

When we applied the threshold RBC indices from the fourth revision of the screening programme, a subset of 187 of the 407 STM subjects (46%) were classified as LRTM (MCV >75 fl, MCH >26 pg, HbA2 <3.2%, and Hb F <3%). This LRTM group included two cases of β-thalassaemia minor, which would have gone undetected by the fourth revision of the screening programme. Haematologic indices for the first case were MCV = 79.6 fl, MCH = 27 pg, HbA2 = 2.4%, and Hb F = 1.1%, and for the second case MCV = 81.6 fl, MCH = 26.9 pg, HbA2 = 2.4% and HbF = 0.5%, with β-thalassaemia mutations IVS110 and -110, respectively.

Of the 271 couples in the STM group, there were 29 where both partners had some form of β-globin gene mutation (Table 3). In 23 of these, one partner had been diagnosed previously with thalassaemia minor (in the first step of screening), and the other partner was STM; in the six remaining couples, both partners were STM. Following molecular analysis, there were 10 couples where both partners had β-thalassaemia minor, and 19 couples with haemoglobinopathies (either haemoglobinopathies for both partners or with one partner having the thalassaemia gene). Prenatal diagnosis is indicated only for thalassaemia major and sickle cell syndromes,¹¹ and was therefore suggested for 20 of the 271 marriages.

Table 3.

Predicted haemoglobinopathies in children of couples selected from the STM group according to the RBC indices and Hb electrophoresis.

Predicted haemoglobinopathy in children	Couples^a (N)	Subjects^b with MCV ≥ 75 fl and MCH ≥ 26 pg	Subjects^b with abnormal band in Hb electrophoresis	Indication for prenatal diagnosis
Sickle-β-thalassaemia	9	5	Yes	Yes
Sickle cell disease	1	1	Yes	Yes
Thalassaemia major	10	1	No	Yes
HbD-β-thalassaemia	4	4	Yes	No
Homozygote D disease	3	3	Yes	No
Other Hb disorders	2	2	Yes	No
Total	29	16

Each couple consists of two partners who both carry β gene mutations.

At least in one partner.

In the STM group, 187 individuals would have been classified as LRTM under the fourth revision of the screening programme. Two of these were confirmed to have thalassaemia minor (1.1% of LRTM versus 4.4% of total STM group; p > 0.05, 95% CI [−0.004,0.026] in the STM group), meaning that 2/18 (11%) of β-thalassaemic individuals in the STM group that would have been detected under the third revision of the screening programme would no longer be detected using the new cut-off values.

Subjects with thalassaemia minor

Among the 984 evaluated subjects, 574 were confirmed to have β-thalassaemia minor; these included 382 subjects in thalassaemia minor group, 135 subjects with a high HbA2 level, and 18 subjects with normal HbA2 level in the STM group, as well as 36 subjects from the personal request group. In six subjects, MCH was greater than 26 pg (with two subjects above 27 pg), and in seven subjects MCV was above 80 fl. One subject, who was known to have family with thalassaemia minor, had MCH and MCV of over 27 pg and 80 fl, respectively (with increased HbA2 levels). Overall, 4.2% of the cases with thalassaemia minor had HbA2 levels lower than 3.5%, and 95.6% of β-thalassaemia minor cases had MCVs lower than 75 fl.

Parents of patients with thalassaemia major

From 218 available cases, there was one subject (mother) with normal RBC indices (silent carrier), and one parent with an MCH of 26.1 pg and MCV of 79.8 fl. Both of these cases had normal HbA2 and Hb F levels, and would not have been detected by the fourth revision of the screening programme; one would be detected by the third revision of the programme. Electrophoresis results from 218 cases showed 14 subjects (6.4%) with normal Hb A2 levels below 3.5%. Two of these 14 parents had MCVs of greater than 70 fl.

Discussion

β-Thalassaemia major is a preventable non-communicable hereditary disease, and given the high prevalence of thalassaemia in certain areas, and the lack of a certain cure (excepting bone marrow transplant), screening has been implemented in many countries.^12–14

Most thalassaemia minor cases have HbA2 levels higher than 3.5%. In the Isfahan thalassaemia screening programme, only 3.3% of thalassaemia minor cases had HbA2 lower than 3.5%.² In our study the prevalence of thalassaemia minor with low HbA2 levels was 4.2% in thalassaemia minor subjects referred to genetic laboratory, and 6.4% in parents of children with thalassaemia major or intermedia. Other studies support these findings, reporting normal HBA2 levels in only 1–10% of thalassaemia minor subjects.¹⁵ This suggests that the majority of children with thalassaemia major could be detected without expensive molecular tests. Some previous studies confirm the use of RBC indices and HbA2 levels for the first step of thalassaemia screening.^7–9 Nevertheless, when molecular examinations were implemented to detect the small percentage of thalassaemia minor people in the STM group in this study (people with low RBC indices but normal HbA2 levels), β-thalassaemia was detected in 4.4% of cases.

The changes to the threshold RBC indices used to determine which individuals should undergo molecular investigation in the fourth revision of the thalassaemia screening programme would result in a seemingly small reduction in diagnosis of thalassaemia for the STM group studied here, from 18 to 16 detections in 407 subjects. Under the fourth revision, 11% of cases with thalassaemia minor in the STM group we studied would not have been detected. Based on previous studies in this province among people referred to counselling centres, 1 of 159 marriages (4412 from 703,082 couples) were referred to the genetic diagnosis laboratory because of suspected thalassaemia.² In the population studied here prenatal diagnosis was suggested after molecular tests for 20 of 271 (1/14) couples in the STM group. Potentially, therefore, if molecular tests were not performed at all, 1 in 2154 marriages would have undiagnosed thalassaemia. If, as recommended under the fourth revision, molecular tests are performed for only those in the STM group not deemed LRTM, we would predict that one out of every 21,545 marriages would have undiagnosed thalassaemia. If the goal is to identify all couples at risk of having a child with thalassaemia major, this rate of failure to detect is of concern.

The thalassaemia prevention programme is performed through two separate steps. The first involves an initial diagnosis of, as well as the identification of STM couples. The second step involves the detection of thalassaemia minor in the STM group. When choosing the optimal RBC cut-offs for screening, the steps of the programme being targeted and how these steps are affected must be specified. If there are no facilities for molecular diagnosis, screening for β-thalassaemia trait can be achieved using MCV, MCH, HbA2 levels and Hb electrophoresis. Based on the results from our group of parents of patients with thalassaemia major, this type of screening would lead to a minimum 93.6% reduction of thalassaemia major in our study. In this step, appropriate cut-off points for RBC indices are only essential for deciding whether to perform HbA2 analysis and Hb electrophoresis. HbA2 analysis is simple and low-cost, so reducing the cut-off points does not meet the economic goals of the revised thalassaemia screening programme. The use of these assays as the first step of the thalassaemia screening programme has been reported elsewhere.¹⁶ The other step is to identify subjects with β-thalassaemia minor among the STM group that constitute about 4.4% of this population. To address the high cost and time-consuming nature of molecular tests, changing the cut-offs can reduce the number of people referred to the laboratory, but 11% of thalassaemia minor subjects in the STM group cannot be identified.

Conclusion

The thalassaemia prevention programme comprises two separate steps. The first step involves the screening of couples to detect (i) carriers and (ii) couples in the STM group. The second step is screening for β-thalassaemia minor in the STM group. Our results show that changing the previously established cut-off points for identifying STM individuals is not justified financially or socially, as it increases the risk of thalassaemia incidence in the second step of the thalassaemia screening programme (screening for thalassaemia in the final suspected subjects).

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Samavat

Modell

. Iranian national thalassaemia screening programme. BMJ 2004; 329: 1134–1137.

Zeinalian

Nobari

Moafi

. Two decades of pre-marital screening for beta-thalassemia in central Iran. J Community Genet 2013; 4: 517–522.

Haghpanah

Nasirabadi

Rahimi

. Sociocultural challenges of beta-thalassaemia major birth in carriers of beta-thalassaemia in Iran. J Med Screen 2012; 19(3): 109–111.

Miri-Moghaddam

Naderi

Izadi

. Causes of new cases of major thalassemia in Sistan and Balouchistan province in South-East of Iran. Iran J Public Health 2012; 41: 67–71.

Ahmadnezhad

Sepehrvand

Jahani

. Evaluation and cost analysis of national health policy of thalassaemia screening in west-Azerbaijan province of Iran. Int J Prev Med 2012; 3: 687–692.

Kalokairinou

. The experience of beta-thalassaemia and its prevention in Cyprus. Med Law 2007; 26: 291–307.

Pranpanus

Sirichotiyakul

Srisupundit

. Sensitivity and specificity of mean corpuscular hemoglobin (MCH): for screening alpha-thalassemia-1 trait and beta-thalassemia trait. J Med Assoc Thai 2009; 92: 739–743.

Bencaiova

Burkhardt

Krafft

. Screening for beta-thalassaemia trait in anaemic pregnant women. Gynecol Obstet Invest 2006; 62: 20–27.

Al-Suliman

. Prevalence of beta-thalassemia trait in premarital screening in Al-Hassa, Saudi Arabia. Ann Saudi Med 2006; 26: 14–16.

10.

Mamtani

Jawahirani

Rughwani

. Value of mean corpuscular volume and mean corpuscular haemoglobin in screening for beta-thalassaemia trait. Acta Haematol 2006; 116: 223–225.

11.

Ryan

Bain

Worthington

. Significant haemoglobinopathies: guidelines for screening and diagnosis. Br J Haematol 2010; 149: 35–49.

12.

Penman BS, Gupta S and Weatherall DJ. Epistasis and the sensitivity of phenotypic screens for beta thalassaemia. Br J Haematol 2015; 169: 117–128.

13.

Fucharoen

Sanchaisuriya

Sae-ung

. A simplified screening strategy for thalassaemia and haemoglobin E in rural communities in south-east Asia. Bull World Health Organ 2004; 82: 364–372.

14.

Shariq

Moiz

Zaidi

. Pre-marital screening for beta thalassaemia in Pakistan: an insight. J Med Screen 2015; 21: 163–164.

15.

Beigom

Rashidbaghan

Maziji

. National approach to premarital diagnosis of trait thalassemia and silent carriers. Int J Clin Med 2013; 4: 91–95.

16.

Weatherall

Letsky

Genetic haematological disorders. In: Wald

Leck

(eds). Antenatal and neonatal screening, Oxford: Oxford University Press, 2000, pp. 243– 243.